Re: Photocopying money

>It's the same chip that's used in vending machines to scan dollar bills.

To my knowledge, vending machines don't do optical recognition.  They have
a pair of photo-sensors, it's true, but these are used in conjunction with
tractors only to measure the bill.  Another sensor decodes the magnetic ink
to determine validity and denomination.

I saw this demonstrated recently on a PBS science program.


Scott Collins         | "Few people realize what tremendous power there
                      |  is in one of these things."     -- Willy Wonka
......................|................................................
BUSINESS.   voice:408.862.0540  fax:974.6094   collins@newton.apple.com
Apple Computer, Inc.   5 Infinite Loop, MS 305-2B   Cupertino, CA 95014
.......................................................................
PERSONAL.   voice/fax:408.257.1746    1024:669687   catalyst@netcom.com

Scott writes:

>To my knowledge, vending machines don't do optical recognition. 

In Japan they do.  Some banks recently got scammed by people putting
photocopied Y10,000 ($100) bills into changers at the banks (getting
10 Y1,000 in return).  They've now done a ROM upgrade on the machines.

A cultural point: Japan is a very cash-oriented society, so nobody
blinked twice when someone went up to the changing machine and plunked
10 or 20 bills into it.

"Robert J. Woodhead" <trebor@foretune.co.jp> wrote:

In Japan they do.  Some banks recently got scammed by people putting
photocopied Y10,000 ($100) bills into changers at the banks (getting
10 Y1,000 in return).  They've now done a ROM upgrade on the machines.

Hmm... I wonder how many free cokes I could get with that method... hehe :)

There used to be a vending machine here where you could push 2 different
buttons at the same time and get two sodas for the price of one.  When I
lived in California, there was a pay phone where you could dial an 800
number and then dial another number and it would charge you as if you
were still connected to the 800 number (free calls!!!)  Oh man... some
vending machine companies overlook the most obvious things...  I guess
it can happen to anybody (remember identify #9 and get 1000000 ep? :)

Hmm... what does this have to do with cryptography?

(This is nostalgiaware - skip past it if you're only into messages
about encryption)

Matthew writes:

>Oh man... some
>vending machine companies overlook the most obvious things...  I guess
>it can happen to anybody (remember identify #9 and get 1000000 ep? :)

This is a reference to my game, Wizardry, an RPG that allowed each
character to have 8 items in inventory.  Alas, the code that checked
for valid keypresses in the "identify item" section had an error:

	IF (ch>='1') _or_ (ch<='8') then id_item(player,ord(ch)-ord('0'));

This allowed any key to be pressed, and ended up twiddling bits in
the data structure.  Pressing 9 gave you extra experience points.
Someone once sent me a list of what _all_ the keys on an Apple II
did.

We deliberately left this bug in the PC version of the game as it had
become part of the legend of the game.

Not much to do with encryption, but Wizardry encrypted it's game strings
with a simple system that nobody broke (I guess tracing p-code was
too much trouble).  It added X*the character position and Y*the line
number+Z*the previous character, as I recall...

"Robert J. Woodhead" <trebor@foretune.co.jp> wrote:

> Not much to do with encryption, but Wizardry encrypted it's
> game strings with a simple system that nobody broke (I guess
> tracing p-code was too much trouble).  It added X*the character
> position and Y*the line number+Z*the previous character, as I recall...

Yeah, that worked pretty well, considering that most people just looked
for text by scanning it with a sector editor.  Encrypting the text made
it undistinguishable from the program code.  Tracing the program code
would have been difficult, considering that it was written in the
original Apple pascal, which used a token compiler (I think).  But,
alas, the character files were in plaintext, and numerous people figured
out how to edit them.  So were the items and monsters.  One day when I
was very bored, I managed to decipher the maze data, and was then able
to build a map of the 10th level where dumapic didn't work.  As I
recall, it was just 2 bits to define each section of wall, arranged in
four groups of data defining each of the four directions, followed by
another block of data which used nybbles to define what was found in the
blocks created by the intersections of the walls.

One thing that I was able to break was the code book that was the
copy-protection for Return of Werdna.  As I recall, it was calculated in
base 9 (or some sort of nine number system which used the digits 1-9 but
excluded zero).  Once I realized that, breaking the code was just a
matter of simple arithmetic.  It was an interesting excercise in
code-breaking tho, and I did it without disassembling the program.
I must say that this was preferrable to the copy-protection in the
original Wizardry, in which my master disk eventually wore out from
overuse and developed bad sectors.  Gee...All this nostalgia has made me
want to play again.  It's a fun game, and the 3-D system was pretty
innovative for the time.  Does Sir-Tech still replace bad disks (from
the Apple II version)?